Trial Heats Do Not a Winner Make

WITH so many voters still undecided about their Nov. 3 presidential choice, and feeling themselves tugged in opposite directions, poll trial heats are of only limited utility. They can't measure a decision that is yet to be made.

Those wanting to get a clearer reading of how the presidential vote may finally go, then, need to look to other sorts of survey questions that seek to tease out inclinations indirectly. One interesting approach here is that used by Yankelovich Clancy Shulman in a recent survey for Time magazine and the Cable News Network. Respondents were asked to assess both President George Bush and Democratic challenger Bill Clinton on a series of dimensions - see the table above - and the results were compared to tho se from a Bush-Dukakis assessment taken at the same point in the 1988 campaign.

We see that while Governor Clinton's trial-heat numbers are higher than Michael Dukakis's were at this time four years ago, he stands in roughly the same position as Mr. Dukakis did in this set of comparisons with Mr. Bush.

Though results vary somewhat from survey to survey, both Bush and Clinton have areas where they are generally seen as stronger than their opponent. The Democrat is preferred in those areas of service delivery, such as environmental protection, where his party has long been favored, and on achieving economic change. The Republican is seen as stronger in foreign affairs, and he personally generates more confidence.

How the electorate finally resolves these contending judgments will determine the Nov. 3 victor.